SURFACE AND ATMOSPHERE OF THE SUN — LINDBLAD 177 



hydrogen which have often an enormous extension. When the spec- 

 troheliograms are extended over the limb of the sun, we find that 

 these filaments are continued as bright prominences outside of the 

 limb. The development of the prominences have been studied with 

 great advantage cinematographically, especially by B. Lyot, R. 

 McMath, and D. H. Menzel, with a great acceleration of the time scale. 

 The mechanism underlying the formation and decay of prominences 

 ofi'ers a great many questions which are as yet unsolved. 



The study of the solar corona was earlier confined to the total 

 eclipses of the sun. The French astronomer B. Lyot, of Meudon, has 

 constructed an instrument, "the coronagraph," by which the inner 

 parts of the corona may be photographed in full daylight. The spec- 

 trum of the corona is mainly continuous, owing to a reflection of sun- 

 light by free electrons, but shows also certain emission lines. The 

 identification of these lines has been singularly difficult. At last the 

 Swedish physicist B. Edlen solved the riddle, and showed that the 

 lines were due to extremely highly ionized atoms. The intense green 

 line is due to iron that has lost 13 electrons. Tlie corresponding tem- 

 perature of the corona is extremely high, several hundred thousand 

 degrees. Several theories have been advanced to explain why the 

 corona has such a high temperature, but there are still many questions 

 in this connection that have not 3- et been answered. 



Though the prominences, as well as the corona, may now be studied 

 in full daylight, the total eclipses of the sun still play a great part in 

 the study of the physics of the sun. Moreover, an exact timing of 

 the occurrence of an eclipse on various places along the eclipse track 

 may give important information about the exact shape of our own 

 globe. 



Ancient eclipses recorded in the countries about the Mediterranean 

 have proved to be of importance in answering the question of the 

 secular increase in the time of revolution of the earth. We may re- 

 construct by modern theory the track of the eclipse, on the assump- 

 tion that our unit of time, the sidereal revolution of the earth, is 

 unchanged. The difference between the old observations and the 

 modern computations has enabled astronomers to draw the conclusion 

 that there is an increase in the time of revolution of the earth of the 

 order of one-thousandth of a second per century. The change is due 

 to the friction of the tidal wave in the ocean, especially in shallow 

 waters close to certain coast lines. 



To the astrophysicist the beginning and end of a total eclipse are 

 of special interest. When the last glimpse of the intensity-radiating 

 surface of the sun vanishes, the spectrum changes character, and 

 instead of the absorption spectrum of the solar disk we get the emis- 

 sion spectrum of the chromosphere. This so-called "flash spectrum" 



